Fixed now, was a thing for 2-3 days

If you have some js knowledge you could set it using f.e.: https://addons.mozilla.org/de/firefox/...entmonkey/
Feel inspired by f.e. https://greasyfork.org/en/scripts/4098...downloader

Rly? I dont see it at all in the search, has it already been fixed? I do only see the replaced version https://animetosho.org/view/n1522817 but for 1/5

For the search term "BD" https://animetosho.org/view/dark-kobay...s.n1521863 seems to be sticking to the top despite being older

It's just a two click option, why fret you four-eyed!

I wish i could not have my theme change every time the new one comes out, rio theme is the most superior one atm.

In case that feature wont come - for things like that (well not torrents but just tried those too) I do use: https://addons.mozilla.org/de/firefox/...nk-gopher/ - open page with all the links you want, may filter like in this case "torrent" and you have all links ready to c&p

Anyway, just found out page breaking is still a thing, on search result pages with resolution being half the 1920x1080.
Example page: https://animetosho.org/search?q=subsplease
On half the 1366x768, backgrounds do not move, although misplaced text still stretches the page horizontally.
Renders: https://imgur.com/a/7l6ziMR
Other pages appear to be fine.

Can you add "Copy to clipboard all magnet links / torrent URLs of torrents listed in current search" buttons to the search page? So I can filter down single-file releases and batch-import that in qBittorrent.

[spam]

[spam]

yeah, people smart enough to get rich should just kill themselves.

> because freaking tvdb decided to go paid

Completely unrelated, but fuck tvdb. Used community input to populate so much data, then went paid. Bunch of Arnolds.

Yeah with what I said I meant that the app is not specifically made for animetosho, I'm just using animetosho database dumps as a source but the app itself just scans text files and matches regex(the database dumps are downloaded using a cron job so it's external to the app) so other sources can be used(like a custom file used to manually add new names). I had thought of using the feed to get new names but since I wanted the app to have a general purpose I opted to just use text files(sadly I still have to rely on xem because freaking tvdb decided to go paid and I need to get the number of seasons and episodes from somewhere).

Not quite sure what sort of response you were looking for, but as for thoughts:

- size is actually smaller, due to compression. ~40KB is probably closer to the typical response size
- serving back the feed isn't really bandwidth limited - it's likely more DB/CPU bound (in fact, compression makes it more CPU bound), particularly since I like using weak servers
- your calculation does assume the system is on 24/7, which will be true for some, but I'd expect many to also turn off the system when not in use
- a lot of people querying data infrequently isn't really a DDoS attack. DDoS attacks generally cause massive spikes above typical load, may target requests to maximise resource drain on the server (e.g. submitting searches with random data to ensure no caching) etc. High load from regular usage will persist over time / not be some one off event, and can often be mitigated using caching strategies (since many requests will be similar) etc.
- in Jan 2021, there was typically around 1.2M reqs/day to feed.animetosho.org. Most of the daily outgoing bandwidth, on that server, is serving those feed requests. Someone doing 48 reqs/day is nothing. Even an additional 10k people doing it doesn't change the picture that much. Worth pointing out that most of the requests are probably from automation software issuing searches over the API, not feed requests.

If you're focusing on updates, the feed API may be more efficient for updates (start off with the dump, update with the feed). For example, Erai-raws updates. More complex filtering is available, but you won't get regex here (you'll need to do that filtering on your end).

But if you're happy with your approach, feel free to continue with it.

Really appreciate your finding these and pointing them out. Yeah, they're definitely problems, but it's at least usable for now - I didn't want to spend too much time trying to patch up an old outdated code base, but if you know of fixes, I can apply them. Otherwise, I can look into them in the future, if I get around to it.

Just seeking confirmation on the other side? Perhaps one thinks it's all good, but they notice it doesn't work, or maybe there's some miscommunication somewhere, etc.

I was imagining a potential torrent client feature: for lots of torrents you obtain via RSS, their time of appearance is usually highly predictable, down to minutes. The idea is to dynamically adjust frequency of / reschedule RSS requests based on probability of obtaining needed torrent, possibly down to one request per week (hereby assuming one-torrent-per-week). The default interval of RSS downloads in qBittorrent (and your ttl) is 30 minutes; given the size of one feed being ~200 KiB, this saves you (assuming 10 and 1 requests per week) ~63.7-65.4 MiB / week / RSS leecher, (* 30/7) ~272.9-280.4 MiB / month / RSS leecher. On the other hand, each torrent release event, even with the requests dispersed over ttl would look like a DDoS attack. As a torrents-via-RSS site owner, what are your thoughts?

Thanks, it no longer breaks down! Although Yuu's still horizontally static and the first table's content takes up more space than the rest of the page. https://imgur.com/a/vqY27aI

> Can I ask an idle question from a guy who is anybody but a web developer? Why do you have a need to ask me whether it's fixed? I figured all recent mainstream desktop browsers would likely produce the same rendering results if given the same render area size (683x654).

... that's what I wanted to write, but then I noticed the theme combo box is rendered overstepping several pixels in Firefox 99 while rendering well in 91.8.0esr, Chromium 101 and Firefox Android 100 Beta. https://imgur.com/a/wm0uOwg  (window size does not matter)

Also, there's a mobile glitch which makes scrolling a bit annoying; if zoomed, Yuu's picture has "mobile auto-hide menubar" moving semantics: when scrolling down, the picture behaves as if it was bolted down to the web page as other content, but when scrolling is upwards it instantly reappears, as if image's bottom was glued to screen's top while scrolling down.

Currently it's only me(I mainly did it to match erai-raws releases since they follow anidb season structure with split cours) but it's made so I can edit a settings file to add multiple files to scan(by setting a file route, a regex and the anidbid and relese name regex groups for each) and multiple release groups to match(by adding a regex and a regex group). You can assign whichever release groups you want to each file and if the release group matches with any release name of the file it's automatically added to the release names list, replaces xem release names for that show and replaces xem's scene mappings with the app own mappings to anidb(using anime-list.xml) since xem anidb's mappings can't be trusted.

Ah I see, that only happens at specific widths. Just needs some adjustment with the media query then - is it better now?

I don't manage the Discord channel, so can't help you there - you can try posting the issue there (if that doesn't time out).

That's an interesting sounding app!
If multiple users are using this, all of them downloading the database daily sounds a little inefficient, though I don't mind it at my end (unless it ends up chewing all the bandwidth).
I can't provide any times. The dumping process is initiated at a constant time (unless I change it), but when it completes can vary, since it does it dumps slowly and with low priority to avoid exhausting resources.

Ok. Here's an example:
The page: https://animetosho.org/view/heroine-ta...l.n1521315
Rendering: https://imgur.com/a/DMfCbLS  (Yuu is me :))
That's not on mobile, through the same thing happens there; that's an old 1366x768 notebook, window taking half the screen space while snapped to the side.
P. S. First time I tried to file the issue on Discord, my message got erased and I got "Timed-out" for "Posting walls of text". That should probably be fixed as well.

So I've developed an app that uses animetosho database dumps to build a list of release names that check a regex and replace xem whenever a show that was found is requested, for this I'm downloading the database once a day(the torrents table), do you have any problem with that? would you prefer if I downloaded it with less frequency? also at what time does the daily dump normally is updated?

Unfortunately, I don't really use Discord, so you should post whatever it is on this page if you want me to see it.

Somewhere around then.

(filed a ticket on Discord)

Oh damnnnn I don't have a mouse

have you tried a mouse-hover on the word "beta" on homepage --a message awaits.

You can't just use the colour options provided?

I really enjoy it when a new colour pops up -Purple-desu!- But I always change it back to Rio lol

Admins - Are the colour schemes still timed to change with the anime seasons btw?

An important moot point against the use of dark mode for better visibility is that the light on dark theme (especially for text) is not necessarily good for everyone. Like I said, when you look at the bright mode, it turns bright. For example entering the multiuporg with bright mode. Such a surprise is not good for the eyes. Especially those who have Astigmatism.

Generally dark pages are better for the eyes, except outside on daylight. Check f.e.  https://rxoptical.com/eye-health/is-da...your-eyes/ f.e. Though I still do use dark mode all time, but on phone many things are colored anyway. As for light themes simply use https://addons.mozilla.org/de/firefox/...arkreader/ - works pretty good.

I hope the next season doesn't use a dark theme. Although it saves electricity, it is not good for the eyes because this site contains third parties and these sites use a light theme. Nor is it an article site.

I very much love this Yuu theme, thank you admin.

Yeah, this website wasn't designed with mobile in mind. It works, but not great there.
I'm not sure what you mean by "breaks the layout". Could you perhaps post a screenshot to clarify?

There's no tool to search through feedback comments unfortunately.

Some pages — the main page, a torrent page — if given small enough horizontal screen space, they are made eligible to be scrolled horizontally by the browser. Doing so breaks the layout. (Firefox 99, Chromium 100)
Is there a tool to search through the feedback section?

Wiki is good as a overview/for features and such but when it gets to definitions in such cases it's hardly of use because the wording counts. And I don't want to check x sites to build some common ground for some archive definition.
It seems the time on the info button I somehow didn't check to open it with winrar instead. (cause custom win theme and symbols they look similar). So it's just the source code (and winrar claiming as archive). xD
Yeah Windows does show simply what's defined in registry (HKCU or HKLM\Software\Classes - sadly rarely still in HKCR).

A tip if you want to know the meaning behind words hunt for specifications/documentations not some random sites. The Info button just opens windows properties box. Now windows can only have one definition active for one 3-character extension. So you get to see that. It doesn't claim that it's an archive because it should do real time scan of file to claim anything. What it is (windows) doing is showing you the one meaning which is active at the time you are looking at it. Glad you were able to better understand the workings. <3

Okay, well I knew the file would be stored 1:1 so that direct play is irrelevant (data could be at the end). I thought the metadata would be stored too. But it rly seems to be dynamically read of the stored data (took a short look at the source - SplitHandler)

Then again you can't use the file as a whole as originally intended (depending on content maybe not at all - videos depend on if the player does allow that). And the question might be if it doesn't match or if the definition is not enough and useage of filesystems and dynamic metadata still should be considered. Since many sites do claim it as archive and 7zip themself - in the Gui (see Info Button) as the source code where the split/merge happens I'll probably stay with that. But I do understand that it's a critical task of definition I wont dare to force in that state.

No they aren't considered archive files. Read the first line under https://peazip.github.io/file-split.ht...er_utility and also the first line of the next paragraph. In peazip's words what you got from here (animetosho) is binary data. In peazip's words merging it will give you larger file. If it were archive then they would have written so.

Now technically open these files from here in hex viewer, what you will find is no metadata which says hello I am an archive. All archive files by standard have to say hello I am an archive in their metadata.

Simple task drop this .001 file in media player. What does it do? It plays it because it's media decoders understand it's a media file not an archive. They don't have support for playing files directly from archives until they were built with that support.

I feel one might not be satisfied with the the prev. answer. So, you may never thought of it as such as it's not for backup-storage purpose or such. And the broad use of the extension with different formats is also confusing. Let me post some links. Read those and the wiki again carefully and I do hope this clears up.
https://file.org/extension/001 - https://fileinfo.com/extension/001 - https://www.file-extension.org/extensions/001 - https://documentation.help/WinRAR/HELP...olumes.htm - https://peazip.github.io/file-split.html (The 7zip documentation else seems to lack info on that part, but already explained what the tool does show)
(@admin about the aim well sry but you've been the one claiming it's not an archive)
How you take it on - mulitvolume/split archives - file spanning does happen in archives - no matter the format are considered archive files.

But indeed it does. This already dates back to the old days of floppys,cds and such. Splitting to parts for storage or in this case transmission is part of that. Don't take Wiki wording for 100% - the compression part is not required to make an archive. But indeed metadata is included (Filename, size, date, .. dunno what else).
Lets shortcut. Open/select a 00x file slipt in 7zip (right click, split file). Click Info Button: 001...Archive.

Who knows... it may even never end!

I'm not sure what your aim is with arguing your point, but assuming you're not just trying to nitpick, an archive feature does not make an archive.

In particular, from the first sentence from your link:

What I'm trying to say here is that split files generated by AT contain no metadata - they're literally the file broken into pieces with no additional packaging.

However, if considering split files as archives is easier to contemplate, feel free to think of them as such.

??? It's still considered and archive NO MATTER the tool used, file extension, .....
And I should have been clear - I meant AT's splits don't offer these functions.
Could you .... just stop.

It actually isn't a split-archive. These end in .mkv.001. There's a separate split file function in 7zip. I didn't know about that since I never used it. If you use GUI then the output filename is like that. The split-archives output as .7z.001. All archive features like compression and hashes are doable in split-archives from my experience.

I know it's a bit ugly, but with different languages and such the best working method I could manage. Also sucks that there is no vale option. (regex (by default/in jd) only can reuse the result of a group (?1) but not pattern)

https://en.wikipedia.org/wiki/Archive_...e#Features - Split Archives may not offer features of a typical archive file like compression or hashes but splitting files is an archive feature nonetheless.

And that's the point, I'm mixing between hosts and don't want to always rename them. Then again I did take a look into jds packagizer and did manage to create s.t., if one is interested: https://www27.zippyshare.com/v/ZcTWkwHr/file.html (cleartext file, can be viewed in advance with npp and json viewer plugin f.e., import in packagizer settings)

When beta testing gonna end ?